Submaximal fitness tests: how useful are they?

As more research becomes available on submaximal fitness testing, we wanted to get an insight from an experienced practitioner and researcher in this area as to how these can be implemented in the field. Tannath Scott of Netball Australia answers six questions on implementing submaximal fitness testing.

In elite sport it’s often difficult to carve out time to run maximal tests to get an understanding of an athlete’s response to training and competition. Because of this and the work of you and others, sub-maximal fitness tests have grown in popularity. With that in mind, what testing options do practitioners have, and which ones are most likely to give us solid data?

The reality for a lot of practitioners is that prioritising time to run maximal fitness tests can be difficult, particularly during the competitive season. While these tests can offer important information on an athlete’s ‘”fitness’”capacities, gathering this information may not outweigh other priorities of that time, the fatigue associated with the testing protocol or a myriad of other factors. As a result, sub-maximal fitness tests (SMFT) have become an increasingly popular tool to assess these or similar capacities in a non-exhaustive manner. While I’m still an advocate for implementing maximal running tests at specific periods (i.e., to prescribe running conditioning, assess the effectiveness of training blocks, profile athlete physical/physiological characteristics), utilising SMFT can allow for a regular flow of information that can assist the management of physical programming.

When looking to implement SMFT, the primary classifications practitioners need to consider are exercise regimen (continuous or intermittent) and manipulation of exercise intensity (fixed, incremental, or variable) (Figure 1) ^[1].

Importantly, the premise of all these SMFT remains the same: identify an altered physiological (performance) state that may indicate a training effect. Currently, there is insufficient evidence to conclude if protocol selection has a meaningful difference on the primary outcome measures on a meta level ^[1]. Therefore, the choice of SMFT is best directed by applying theoretical frameworks and beliefs, and protocol measurement properties, alongside considerations of your environment and constraints.

For example, practitioners in field and court sports may select a running-based test over a cycle test, due to a greater level of transfer to their environment. However, in regions that experience greater fluctuations in environmental conditions across a year (heat, humidity, rainfall), this can cause greater noise in the outcome measures.

As a result, practitioners may need to take a more conservative approach to inferring a “real” change in the physiological state of individual athletes.

Another example may be the reluctance of a staff to provide opportunities for structured testing within the season. As such, practitioners may look to include higher intensity standardised runs (~50-60m) ^[2,3] or drill- or game-based activities ^[4-6]. However, these may present greater variability in the athletes’ movements and physiological or neuromuscular response, potentially limiting the sensitivity of testing.

Working primarily in running-based team sports, I lean towards running the athletes during testing. However, whilst this may be more reflective of their performance environment, there is a trade off as you introduce external elements that may contribute to noise. When implementing an SMFT, practitioners should consider the outcome measures (cardio-respiratory/metabolic, subjective or mechanical) ^[7], and purposes of testing (monitoring acute versus chronic, positive versus negative effects) ^{[8, 9]}.

What are the most actionable metrics that coaches should be collecting from a sub-max test?

SMFTs have been most commonly implemented to assess the internal response to a standardised external dose. Here, heart rate-based indices observed throughout and following the test are the most common surrogate measures of physiological function. These include:

Exercise Heart Rate (HRex): This is the most commonly used and reported measure in SMFT, due to its strong relationship to oxygen uptake during continuous, steady-state exercise. HRex is typically calculated as the average heart rate in the last 30-60 seconds of the protocol. HRex best represents the cardiovascular fitness of individual athletes. While still somewhat contentious, HRex has also been proposed as a marker of shorter-term physiological stress ^[1].

Heart Rate Recovery (HRR): Heart rate recovery is the decrement in heart rate response in the 10-180 seconds post-test, calculated as the absolute or relative difference between the HRex and the post-exercise HR . A longer period (e.g., 120 seconds) appears more reliable that a shorter timeframe (e.g., 60 seconds) ^[10]. HRR may be collected while the athlete is lying down, sitting, standing or walking, though limiting movement is best. Heart rate recovery is characterised by parasympathetic reactivation and sympathetic withdrawal in response to the cessation of exercise ^[11], reflective of the hemodynamic adjustments to body position ^[12,13].

Heart Rate Variability (HRV): is the variability in the time intervals between consecutive heartbeats and represents the regulation of cardiac autonomic nervous system balance ^[12]. When used in the context of SMFT, these measures are usually analysed 3-5 minutes post-test cessation and calculated as time-related variables.

A reduction in HRex may indicate improved exercise economy, which may in turn indicate the development of aerobic fitness ^[14]. Conversely, an increase in HRex reflects a negative response, such as de-training of the cardio-respiratory system, likely due to central adaptations (i.e., left-ventricular function) ^[15]. It also appears that training stimulus completed in the 72 hours prior to testing can inversely affect on the results (i.e., high training load causing decreases in HRex) ^[7], potentially due to exercise-induced plasma expansion, among other causes ^[12,16,17]. In addition, an increment of HRR or vagal-related HRV measures are considered as positive effects ^[12], reflecting the reactivation of the parasympathetic system and hemodynamic adjustments post-exercise ^[12]. That said, with higher intensities eliciting increased blood acidosis that simulate the metaboreflex, and therefore may reduce HR decay post-exercise and alter HRR and vagal-related HRV results ^[12]. Lastly, other internal measures such as rating of perceived exertion, hold some theoretical value (e.g., if an athlete perceives a SMFT as less stressful they may have experienced positive psychophysiological adaptations ^[18]), though there is currently limited evidence in this area.

The common typical error of HRex is 1-2%, so a change of ~3% may be considered a “real” change to an individual’s physiological state ^[1].

We found that a decrement of 5.6% from an individualised SMFT correlated with a change in the 30-15 Intermittent Fitness Test (1 stage; 0.5 km·h^-1) ^[7]. Our preliminary analysis showed -5.6% in HRex also equates to a ~30 second improvement in a maximal incremental running time trial and an improvement of ~3 – 4 ml·kg·min^-1 V̇O₂ max.

Through understanding the measurement properties of the testing protocol relative to our population, we are able to assess the adaptations of our athletes over training blocks in similar fashion to a maximal running test (see Figure 2).

When first implementing SMFTs, understanding the interrelation of these variables can be difficult due to the complex interactions with training volume, intensity and density. As such, interpret these heart rate indices in context, conduct then regulary (every 2 – 4 weeks), with limited physical activity in the proceeding 72 hours, if feasible.

Is it possible to individualise a sub-max test to get more meaningful results? If so, how would a coach go about that?

The indivdiualisation of SMFTs is an area of player monitoring that has intrigued me for several years. I believe there exists a strong theoretical rationale behind individualising an external dose to better standardise the internal response across the playing group.

In sports like rugby codes, where there is heterogeneity in inter-individual physical qualities and body compositions, having all athletes run a submaximal Yo-Yo or 12 km·h^-1 continuous run seems sub-optimal. This is due to the spread of heart rate responses that may occur. For example, it may not be uncommon to observe one athlete record a HRex of 78% of their HRmax, while another up at 91%. Having these athletes record a difference of 13% points may limit the interpretation of the individual magnitude of change over time. As a result, I favour having athlete baselines in a similar range. I consider a baseline testing at ~80-85% HRmax to be an appropriate response for sub-maximal activity. However, the physiological justification for the most optimal sub-maximal activity (and if it matters) is not well known.

Due to these experiences, we have investigated the reliability of an individualised SMFT derived from the 30-15 Intermittent Fitness Test ^[19], compared to both the submax Yo-Yo and the 12 km/h continuous run. Here we observed improved HRex reliability during SMFT when using individually prescribed sub-maximal shuttles (57–64 m, TE = 1.3 %) versus continuous running (TE = 2.4%) and modified Yo-Yo (TE = 2.6%) protocols at absolute (athlete-independent) running velocities.

This is important, as the greater the test reliability, the more sensitive it may be to detect changes in physiological function (assuming it also possess sufficient convergent validity).

Individualising SMFT also has some considerations that practitioners must be aware of. One of these is that practitioners likely individualise SMFT distance at the beginning of each season, and then keep it constant for the remainder. This creates constraints if practitioners want to look at changes in outcome measures over time, as the external dose may have changed.

Whilst I am still a supporter of individualised SMFT, practitioners should consider if long term monitoring and extra preparation are of greater importance than the possibly improved reliability and sensitivity.

Are there any considerations when running sub-max tests with younger or less experienced athletes?

Due to their non-invasive nature, SMFT offer a great opportunity to assess the physiological state of youth and inexperienced athletes. Current evidence suggests that there is no clear effect of age or level of competition on the reliability and validity of outcome measures during SMFT.

Shushan ^[1] reports an average ~0.12 % points reduction in typical error for every three years increase in age or about 0.3% points between youth and senior (under or over 18 years old, respectively). Whilst these changes may be insignificant and suggest the tests can be used in the same manner across age groups, practitioners may choose to apply a slightly more conservative approach when interpreting findings in youth athletes (e.g., 4 – 5 % change).

As with adult populations, trying to standardise the approach and conditions around your SMFT protocol is key. From my experience, youth athletes may be undertaking additional extra-curricular activities which can often make this more challenging. Regardless, current evidence demonstrates the application of SMFT across a variety of populations.

Is there any merit to collecting biomechanical or external load data during sub-max tests? What do we know about this area?

This is a growing area that holds dangerous value in my perspective. While there is an emerging interest in the utility of these mechanical measures, their measurement properties and underpinning mechanisms are not fully established. Our lab recently assessed the convergent validity of accelerometer-based measures, finding strong associations between decreases in total and vertical accelerometer load and an improvement in maximal running performance (30-15 Intermittent Fitness Test) ^[7]. A reduction in acceleration loads may be due to reduced leg stiffness ^[20-22], morphological adaptations ^{[23, 24]}, or improved gait coordination and running economy ^[25]. Other than reduced leg stiffness, these adaptations appear beneficial to running performance. For example, excessive changes in momentum (and increases in net and total vertical impulse) are wasteful motions, requiring greater metabolic demands ^[25]. Yet due to the complex interaction between kinetic and kinematic performance in the field ^[23], coupled with the lack of concurrent measures of neuromuscular performance and/or criterion measures of morphological and biomechanical functions, drawing definite conclusions is currently difficult.

Having athlete wear GPS units between the scapulae during data collection may not be ideal for precisely capturing lower limb stiffness and movement ^[26]. However, these results should offer some promise that we can meaningfully record mechanical properties during SMFT. There is also a great push to quantify and understand these lower limb kinetics and kinematics, with advances in foot-worn accelerometry, among other field assessment tools.

Collectively, this work may offer another layer of support for using standardised SMFT to observe the multidimensional constructs of internal load, and provide useful information on the athlete performance state.

What are the biggest mistakes you see young practitioners/clinicians make and what advice would you give them to help?

One of the biggest traps we can fall into is the “collect data for a year so we can make decisions next year” mindset. While I understand the premise, I often feel this sentiment is used to justify using an information source without a true understanding of how it can and will effect intervention and decision-making within the performance environment.

Given the plethora of often shallow information available, we may feel pressured into keeping up with the Joneses, ultimately ending with analysis paralysis.

My biggest advice would be to do the simple things well, then build once you have a system that can instigate change. To do this, we need to understand where the biggest areas of growth are for our athletes. Do we capture relevant, valid and reliable data that can assist in providing information on how, why, and when interventions should be programmed?

Ultimately, does the information we collect have the ability to create change within our environment, and not in three years? I don’t think it needs to be more complicated than that.